-Learning and Acting Upon Customer Information: An Empirical Investigation of Service Allocations with Offshore Centers

Ever since the 1990s, the role of call centers has been transformed from a cost to be minimized to a crucial element that performs integrated marketing functions. Call centers have become among the most crucial corporate assets to grow customer relationships and firm profits. In this paper, we formulate the traditional operation management decision of call allocations as a customer relationship management problem in which the firm learns about heterogeneous customer preferences, balances the trade-offs between short-term service costs and long-term customer reactions, and makes optimal service allocations that best match customer preferences. Using customer call history data from a DSL service, we first estimate customer response models to parameterize the relationship among call allocation, service duration, and customer retention. On the basis of the estimation results, we conduct simulations to derive the optimal service allocation decisions. We demonstrate that learning enables the firm to improve its knowledge about customers and make customer-centric allocation decisions. And acting on future marketing consequences allows the firm to make proactive decisions to prevent customers from leaving and improves their long-term contribution to the firm. We show that with continuous learning and acting on long-term marketing consequences, the derived optimal allocation decisions (1) reduce service costs, (2) improve customer retention, and (3) enhance profit.